Recent lawsuits involving alleged chatbot-related harm have intensified public scrutiny around conversational AI and raised new questions about liability, user vulnerability, and juror perception. As AI tools become more integrated into daily life — and increasingly human-like in the way users interact with them — courts and juries may soon be asked to evaluate issues that feel fundamentally different from prior technology litigation.
DOAR Director Ellen Brickman, Ph.D., a nationally recognized jury consultant with extensive experience advising on high-stakes litigation and juror perception research, shares her perspective on how jurors may respond to emotionally charged AI cases, the challenges companies face in defending these claims, and why conversational AI may be viewed differently than other forms of technology.
Q: As lawsuits involving alleged chatbot-related suicides and violent incidents continue to make headlines, what do these cases say about growing public concerns surrounding AI?
Publicity around these cases is likely to increase concern, particularly among those who are less comfortable with AI to begin with. I think for some people, these concerns will also be conflated with growing concerns about data privacy, again concentrated in the sector that is less comfortable and less familiar with these technologies.
Q: Several recent lawsuits involve claims that users formed emotional relationships with chatbots. Why is that likely to be an important issue in these cases?
From a jury’s perspective, it’s an aspect of human-technology interaction that differs from what people are used to seeing or hearing about. Some jurors will undoubtedly struggle to wrap their minds around this idea, and it will be fascinating to see how they react to this new concept. It also raises broader questions about how people perceive and emotionally engage with AI systems in ways that may feel very real to the user, even when they understand they are interacting with technology.
Q: In cases involving alleged chatbot-related harm, who are people most likely to hold responsible — the user, the company, or the technology itself?
In a legal context, there isn’t likely to be a distinction between the technology and the company that created it. There’s no mechanism for holding technology liable other than to assign liability to the company that owns it. As for how blame is apportioned between the user and the company, it’s likely to vary greatly depending on the circumstances. In a recent DOAR study examining juror attitudes toward chatbot-related harm, a scenario in which the chatbot provided information on fatal poisons elicited more blaming of the company (unsurprisingly) than a scenario in which no such instruction was provided.
Q: Are people beginning to view conversational AI differently than other forms of technology or social media?
I think they are likely to. Social media involves relating to or communicating with other people through a technological platform. With AI, there is no other person; it’s purely a human-computer interaction or relationship, and that’s something most people haven’t experienced before. As these tools become more conversational and personalized, I think people are beginning to engage with them in ways that feel fundamentally different from traditional technology or social media platforms.
Q: Many AI companies warn users not to rely on chatbot advice. How effective are those disclaimers likely to be in front of a jury?
I would assume it would depend on who the user is. I suspect a disclaimer won’t have much effect if the user is a teenager or someone else who is particularly vulnerable. Also, because the chatbots are designed to capture your attention and keep you engaged, jurors may perceive them as inducing reliance. I believe that perception would far outweigh a one-line disclaimer.
Q: What kinds of facts or themes are most likely to influence jurors in emotionally charged AI-related cases?
Issues of vulnerability will be important – if the user is a minor and/or struggling with mental health issues, that will certainly play a role for jurors. What efforts the company made to build in safeguards and prevent harm will also be important – jurors can be influenced by a perception that the company does or does not care about the effects of its products.
Q: What should companies developing or using conversational AI tools be thinking about right now from a litigation and reputational risk standpoint?
Companies should be and already are establishing new protocols to protect minors. They should also stress-test their chatbots and the established protections to see how they perform over time. To date, protections that are effective at the beginning of an interaction (e.g., training the chatbot not to provide any information about suicide methods) tend to degrade over time. These are just a few examples of the kinds of concerns that need to be considered.
Q: Are there similarities between the way AI litigation is developing today and the early waves of litigation involving social media, cryptocurrency, or other emerging technologies?
They all require introducing jurors to what may feel like a new world, especially for older jurors. From a jury perspective, we’re likely to see younger jurors become more dominant forces in deliberations as their older counterparts rely on them for expertise. It can change the way we evaluate each juror’s leadership potential during voir dire.
